Peptide Drugs QSAR Study Based On The Three Different Methods

The purpose of quantitative structure activity relationship(QSAR)is to establish a mathematical model between the molecular structure and biological activity by using the method of chemical combination of mathematics and computer science.The peptide sequence was composed by amino acids,and amino acids play a decisive role in the biological function and structure of peptides.Therefore,the QSAR study of peptides is based on the analysis of amino acids at the molecular level.This paper mainly uses the methods of amino acid descriptor,the comparative molecular field analysis(CoMFA),comparative molecular similarity indices analysis(CoMSIA)and topomer comparative molecular field analysis(Topomer CoMFA)to establish QSAR models and comprehensive study the mechanism of action for different peptides.The structural parameters of the peptides were established by multiple linear regression(MLR)and partial least squares(PLS).The research of this paper includes as follows:In the first part,based on the spatial structure characteristics of 20 kinds of natural amino acids,three kinds of 3D structure index of amino acids were calculated by the software: Gemetrical descriptors,Molecular-propepty descriptors,and Whim descriptors.A new descriptor-SVGMW was extracted from principle component analysis in feature extraction.The structural characterization of angiotensin converting enzyme inhibitors,bitter tasting threshold of di-peptide,and oxytocin were performed by using a SVGMW,then QSAR models were built by MLR.The internal and external double verification methods are used to verify the stability and prediction ability of the model,and obtain satisfactory results.That shows SVGMW covers nearly all the information of Gemetrical descriptors,Molecular-propepty descriptors,and Whim descriptors.Therefore,the descriptor SVGMW has a certain guiding role in the study of QSAR.In the second part,based on the spatial structure characteristics of 20 kinds of natural amino acids,three kinds of 3D structure index of amino acids were calculated by the software: Geometrical descriptors,Eigenvalue-based indices and Randic indices.A new descriptor-SVGER was extracted from principle component analysis in feature extraction.The structural characterization of angiotensin converting enzyme inhibitors,bitter tasting threshold of di-peptide,and oxytocin was performed by using a SVGER,then QSAR models were built by PLS.The internal and external double verification methods are used to verify the stability and prediction ability of the model,and obtain a satisfactory result.The results show that the amino acid descriptor SVGER could characterize the structural information of the peptide drug,and the QSAR models also have better stability and prediction ability.In the third part of the article,the paper uses CoMFA and CoMSIA to study QSAR for bitter tasting threshold of di-peptide,oxytocin and angiotensin converting enzyme inhibitors.The models established by PLS have good prediction ability.In addition,combined with the potential map analysis,the results show that the positive and negative electric properties and the size of the amino acid residues all influence the drug activities.And can provide guidance for the design of peptide drugs.In the fourth part,the paper using the Topomer CoMFA method based on R group to study the quantitative structure-activity relationship between the bitter tasting threshold of di-peptide,oxytocin and bradykinin-potentiating peptides.And it obtains the fitting of the optimal model and complex correlation coefficient,the established mathematical models have good predictive ability for peptide drugs.Based on the model combined with the potential map analysis,the results show that the positive and negative electric properties and the size of the amino acid residues are closely related to the activity of peptide drugs,which provide a theoretical basis for the design of peptide drug.